In this paper a study of descriptions, recommendations and requirements of methods for doing systematic reveiws is presented. The study draws on document material presented by three prominent brokerage agencies working in the field of educationl sience. The data material provides examples of approaches to systematic reviewing that indicate the existence of one common and overall methodology, but also how agency specific approaches and procedures are in play. Findings show that there is procedural variation between agencies in how they proceed in their making of systematic reviews and how varying issues and elemements of the review process are being emphasised by the agencies. In the growing knowledge base on research synthesis variation in approaches to research synthesis is acknowledged but there seem to be little focus on the consequences of variations in method and procedure for the understanding of evidence. The findings call for a discussion on the main differences between agency specific approaches in systematic reviewing and the consequences of such variations for what count as evidence in education science.